Six blood samples were obtained from each of a group of 33 healthy males between the ages of 19 and 31, following which radioimmunoassays were used to determine the serum concentrations of testosterone (Tser), 5 alpha-dihydrotestosterone (DHT), and estradiol (E2). In addition, the free testosterone (Tsal) was also measured using saliva samples provided by 23 of the subjects. A questionnaire of our own design was administered together with the Sixteen Personality Factor Questionnaire (16 PF-Test) at the time of the first blood sample in order to check the long-term stress loads of our subjects as well as their abilities to deal with stress. During the investigational period, subjects kept daily records of their sleeping and working hours and noted the appearance of stressful situations. Weather data for Hamburg was also included as a variable in this study. A number of significant relationships between sex hormones and stress could be ascertained; however, it should be kept in mind that the correlation coefficients are low and explain only a small percentage of the variance between the variables. The stress variables "weather condition" and the "Q4" factor of the 16 PF-Test are significantly related to E2 (intersubject correlations). For all samples of all subjects, psychic stress correlates positively with the ratio of Tsal/Tser. There is a significant positive intersubject relationship between Tsal and long-term plus concurrent somatic stress, while somatic stressors on the day preceding a blood and saliva sample (acute somatic stress) correlate positively with Tsal and Tser.